Start by identifying what is actually failing. Check whether the whole website is unavailable or whether the problem affects a particular page, login, form, checkout or email notification. Note the exact address, error message and time it happened. If possible, compare another device or network so you can distinguish a local connection or cached result from a wider problem.
IN THIS ARTICLE
01 Check your service before changing data
Open the relevant hosting service in your client area and check its status. Review any account message, recent change or usage warning that may explain the issue. If you can access the Account Manager or WordPress Manager, inspect the available logs and activity without immediately deleting files, resetting settings or restoring an old database. Those changes can remove useful evidence or create an additional problem.
02 Investigate with the Engineer or contact support
The relevant AI Engineer can help investigate a specific symptom. A useful request is: 'The checkout page started returning an error after a plugin update this morning. Please inspect the recent errors and propose the smallest suitable repair before changing anything.' Tell it about recent deployments, updates or DNS changes. Review its proposed action and check what recovery method is available before approving a material change.
Contact our support team promptly for a widespread outage, suspected compromise, account suspension, payment problem or anything outside the tools you can access. Include the service reference, affected URL, time and timezone, error text and the changes already attempted. Share only the necessary log excerpts and remove secrets or personal customer data.
03 Protect live business data and verify the recovery
For a live store or booking service, explain the business impact and whether new transactions are still arriving. Avoid restoring or synchronizing a database without checking what newer data would be overwritten. After any repair, verify the failed function and related customer journeys, then monitor for recurrence rather than relying only on a success message from a tool.
What should I check first when a website stops working?
- Check service status, current DNS destination and the recent deployment or update history.
- Collect the relevant error entry and reproduce the smallest failing action before altering files or settings.
- Apply one authorised correction at a time, then repeat the same test and note the result.
If the problem began after an Engineer session, identify the actual operation that completed. Supported file Undo, a WordPress revision and a database restore cover different data. Choose the narrowest suitable recovery method and preserve newer orders or submissions before restoring an older database.